Ep 36: Storytelling as a Critical Human Skill in the Age of AI with Joe Lazer

[00:00:00] Welcome to Elevate Your AIQ, the podcast focused on the AI-powered yet human-centric future of work.

[00:00:05] Are you and your organization prepared? If not, let's get there together. The show is open to

[00:00:09] sponsorships from forward-thinking brands who are fellow advocates for responsible AI literacy

[00:00:13] and AI skills development to help ensure no individuals or organizations are left behind.

[00:00:18] I also facilitate expert panels, interviews, and offer advisory services to help shape

[00:00:23] your responsible AI journey. Go to ElevateYourAIQ.com to find out more.

[00:00:28] Hey everyone, welcome back to the show. In this episode, I had the pleasure of sitting down with

[00:00:43] Joe Lazer, bestselling author of The Storytelling Edge and the creator of a popular newsletter by

[00:00:48] the same name on LinkedIn and recently launching in more depth on Substack. Joe and I dive into the

[00:00:54] ever-changing world of AI, discussing the challenges businesses face in implementing these technologies.

[00:00:59] Joe shares insights from his journey at Contently and then A-Team, where Joe is currently their fractional

[00:01:04] head of content, exploring how companies can effectively use AI without losing the human touch

[00:01:08] in content creation and staying authentic. Beyond that, Joe even gave me some personal

[00:01:12] tips on writing and storytelling that I'll be putting into practice. If you're curious about

[00:01:17] the future of AI and how it intersects with storytelling and creativity, you won't want

[00:01:21] to miss a minute of this conversation. Enjoy it. Hi everyone, welcome to another episode of Elevate

[00:01:28] Your AIQ. I'm your host, Bob Pulver. With me today is Mr. Joe Lazer. How are you doing today, Joe?

[00:01:34] Doing well. You know, you got me here on Halloween, so I'm about to take my kid out right after this and

[00:01:41] get to those mean streets of Park Slope, get some candy, show him off in his first Halloween costume. I'm excited.

[00:01:48] Awesome. Yeah, no, that'll be a lot of fun. So yeah, I appreciate you doing this and I was wondering if you could just give

[00:01:56] my audience a little bit about your background and how you were introduced to AI. I know you've spent a lot of time in this

[00:02:03] space trying to teach others about what's possible with AI and your work at A-Team and stuff like that.

[00:02:09] Yeah, sure. So I'd say that probably my first exposure to AI came, you know, close to a decade ago when I was at

[00:02:16] Contently before I was at A-Team. So Contently was a content marketing technology company and one of our bigger

[00:02:23] clients was IBM Watson, right? So we were working with them on a lot of their content around AI.

[00:02:30] We also ended up integrating IBM Watson into our platform to get a lot of AI-driven content insights,

[00:02:37] which is very early on, right? These are things more like on voice and tone analyzer,

[00:02:41] like here's the voice and tone your audience is most likely to respond to. Here's automated

[00:02:45] recommendations on how to tweak your voice to do that. Stuff that's baked into Grammarly Pro

[00:02:50] now that I think we just gloss over, but was really cool at the time a decade ago, you know,

[00:02:56] we'd show it to prospects and be like, wow, this is the most awesome thing ever.

[00:03:00] And then after when I left Contently, I went to be on the founding team running marketing for this

[00:03:07] future of work, AI startup called A-Team. And the idea behind A-Team was to build an AI team graph that

[00:03:15] would understand who worked well with whom on what in small innovation teams, right? So learning from

[00:03:22] the experience that people have working together, how fast they build, how well they rate their

[00:03:29] experience with one another and do that over and over again. So you can figure out to optimize what is

[00:03:35] the best team to bring in, build any given products. You do it as quickly, as high quality, and as cheaply

[00:03:41] as possible. And I thought that was a really cool vision because it was building out a lot of what

[00:03:46] we'd done at Contently, which is bringing together creative teams, but doing that for product and

[00:03:51] engineering teams, which, you know, people are willing to spend a lot more money on and using AI to kind of

[00:03:57] master what had been, you know, a very manual process for us when we were at Contently.

[00:04:01] And through that, you know, it's also being at this company with a lot of very smart product and

[00:04:08] engineering folks, right? And even before ChatGPT came out, my team was experimenting a lot with

[00:04:16] different generative AI tools. You know, we were playing around with Jasper for ad copy, with mid-journey,

[00:04:23] with runway to figure out, okay, can we create really cool and interesting videos much more cheaply than

[00:04:29] we would otherwise. And then of course, once ChatGPT came out, that created this next wave groundswell that

[00:04:37] both very much changed the way that we went to market at A-Team because we've been doing AI projects. And I was

[00:04:44] like, wow, this is the one thing that everyone's interested in. And I think like with everyone leading

[00:04:50] marketing at the time, there was increased pressure to figure out, okay, how do we really use this for

[00:04:55] ourselves to be, you know, leaders, innovators, whatever, you know, bullshit business buzzword

[00:05:02] you want to use on the topic of AI? Because that was a big pressure, I think, for a lot of marketing

[00:05:07] teams at the time. So first of all, Joe, you and I were basically introduced to, I would say modern AI

[00:05:15] going back decades at almost the same time, but from different perspectives. So I was at IBM,

[00:05:22] working on an IBM research campus where we had the mock Jeopardy studio set up and we first exposed

[00:05:31] clients to what IBM Watson was. So from that and understanding why cloud computing, big data,

[00:05:40] advanced analytics were all necessary to, you know, make this leap with AI. And then I wound up

[00:05:47] trying to sell some of those same solutions. So I remember the partnership with Contently. I was

[00:05:52] there when we formed a Twitter partnership to ingest all that social data and run sales,

[00:05:57] social analytics on it, partnership with a weather company, which they wound up acquiring to get all

[00:06:01] kinds of other, you know, sort of localized, you know, data that affects, you know, retail and all kinds

[00:06:06] of industries. But some of those capabilities that you were describing in terms of how to understand,

[00:06:13] you know, content, how to understand personality and tone and all those kinds of things were,

[00:06:19] I mean, they were just, you know, APIs at the time, right? It wasn't built into the platform and

[00:06:24] it certainly hadn't, they hadn't really built out the ecosystem that Contently was one of the first

[00:06:29] to really start to take advantage of it and play with it and see what was possible. The other sort of

[00:06:35] huge parallel is the last program that I was running at IBM was, I had a lot of similarities to the A-Team

[00:06:44] model, which is now that people know what's possible with AI and what these APIs can do,

[00:06:52] what ideas do you have? How do you vet those ideas? How do you build a team around those ideas

[00:06:57] and move them from an idea to reality? And I say reality instead of commercialization,

[00:07:04] because some of those ideas were internal basing, right? Maybe they're for cost, you know, savings or

[00:07:10] efficiency and things like that. But certainly there were a lot of commercialization, you know,

[00:07:14] opportunities as well. But all the work that I did in idea management, innovation management at IBM,

[00:07:20] and then seeing what AI could do, it's like, okay, well, what if you used AI to help, you know,

[00:07:27] strengthen the idea? What if you use AI to, you know, help you select the right people with the right

[00:07:32] skills to build that team? And then how do you go and use AI to do the market research on the viability

[00:07:39] of the idea and the SWOT analysis? And then how do you go and use AI to target specific potential sponsors

[00:07:46] and investors for those ideas? So it seems like in a lot of ways, a team, you guys just basically,

[00:07:53] you know, commercialized and hyper accelerated, you know, what's what's possible in that sort of

[00:07:58] life cycle? Well, I think these are all the big questions that companies are still grappling with,

[00:08:03] right? It's like, what's the actual high ROI use case that I can have for this technology?

[00:08:08] There's been a little bit of that shiny object syndrome with AI over the last couple of years,

[00:08:13] where it's like, oh, like we can, let's just find some use case for it. There's pressure from the

[00:08:18] board top down for us to figure out what our strategy is. And I think the biggest thing that

[00:08:23] we've found is that with AI, it's really easy to get to like 70, 75% done to have a prototype that

[00:08:31] works well enough in certain contexts to kind of prove out an idea. But then it's very hard to do

[00:08:38] that last mile, that last 20 to 30%, especially when it comes to things like having a scalable

[00:08:43] infrastructure, minimizing hallucinations. So much of the work we're really starting to do now is in

[00:08:51] partnership with, you know, other larger frontier model type companies who have realized that you

[00:08:58] can't really just give most corporate clients all these APIs or just, you know, give them chat GPT

[00:09:06] enterprise and say, go have at it. You're not actually going to get any gains from that. And a

[00:09:12] lot of it, you know, requires having a knowledge of what these technologies can do. And I think one

[00:09:18] of the advantages of working with, you know, fractional specialized folks in this is that

[00:09:22] they've seen it across six to eight companies. They already know what giant pitfalls there are

[00:09:26] along the way. And the second part is that they have that zero to one startup mentality and experience

[00:09:34] for building new products, which if you know, we're being honest, quite frankly, just isn't the

[00:09:38] purview of most IT teams inside of large non-tech organizations to do that. So it's a kind of

[00:09:45] different muscle and cultural mindset that you need to bring in. So, you know, as much as we're working

[00:09:51] on it, I think we're still figuring out right what the best model is for all of this. Like we're

[00:09:56] literally learning every day and with every engagement. Right, right. And so a lot of the folks that come to

[00:10:02] you, you know, to, to a team, I mean, are they, you know, budding entrepreneurs? Are they, are they

[00:10:07] startups who are stuck or are they, you know, companies who just want to, you know, do some

[00:10:13] kind of spinoff in a skunkworks thing that doesn't abide by the same rules and inhibitors to, to

[00:10:19] innovation within their own company?

[00:10:21] It ranges, right? So, you know, we have, we have customers in every market and startup and

[00:10:27] mid-market and enterprise with startups. I think you get two buckets. One is more of an AI native

[00:10:33] startup. They know what they want to build. They have the idea. They just need to move really quickly.

[00:10:38] So they want really good engineering or design or data science talent that they can get up and running

[00:10:44] in the next, you know, week from a team to fill in those holes. Cause they don't want to waste,

[00:10:48] you know, four plus months on a hiring cycle. Yeah. That's what kind of bucket one bucket

[00:10:52] too, is that it's more of a non-tech native SMB or they built their product, you know, like a decade

[00:11:00] ago. They want to bring in someone who can, you know, give them that team to build, say a new AI

[00:11:08] driven LOB to take their business to the next level. Cause they see the writing on the wall that

[00:11:13] if they just stick with their legacy product, it's probably not going to work out well in the long

[00:11:17] run. Then when mid-market and enterprise, it tends to happen much more on the strategy front first.

[00:11:23] So kind of help us figure out what's the right use case here, sell it in internally. And then

[00:11:29] bring in the right type of team with the right type of experience who's done this before to build it

[00:11:35] out. And again, I think that's maybe the most valuable thing right now that really good fractional

[00:11:42] folks in our network bring is the fact that they've just done it across, you know, a half dozen or a

[00:11:48] dozen different companies. They're very good at these very specific stages of the development process.

[00:11:54] They know what's going to go wrong, which if you just have built it at one company before, you just

[00:11:59] don't have as good of a breadth of understanding of what the potential pitfalls are. It's a case where

[00:12:06] I think companies are just getting there where they're ready to move past the prototyping and

[00:12:11] pilot phase and really adopt generative AI. Like all the hype two years ago where it felt like every

[00:12:17] analyst on earth was tripping over themselves to project how much GDP AI would add in the next year,

[00:12:24] how many jobs it would replace, what the pace of transformation would be. I mean, like,

[00:12:28] I think we all kind of knew in the back of our minds that that was likely bullshit and we were

[00:12:34] likely headed for the bit of trough of disillusionment that we're in right now,

[00:12:38] a query and a gardener. But that's a healthy part of the process because now we actually figure out,

[00:12:42] you know, what's the right application of this technology and what works.

[00:12:45] Yeah, no, that makes total sense. And I think you're right. You guys have built a pretty impressive

[00:12:50] network of some of those fractional folks because I know a few of them and you're right. The exposure to

[00:12:55] different companies in different scenarios is really important, not just from maybe a data and analytics

[00:13:02] maturity standpoint. Do you even have the foundation to build something that can be,

[00:13:07] you know, scaled across an existing enterprise, maybe overcoming some of the legacy baggage,

[00:13:15] but also just culturally, you know, every company's different and you may have different,

[00:13:20] you know, approaches to how fast people want to move or how quickly you want to, you know, maybe

[00:13:26] move from experimentation in one domain and move on to others. So just different appetites, I suppose.

[00:13:32] Yeah. And when it comes down to it, like most companies want to do the same thing with Genuine

[00:13:36] of AI right now. They want to be able to talk to their data and get insights out of it. Like when

[00:13:41] it comes down to it, that's, that's what 90% of these ideas or implementations comes down to.

[00:13:47] What I've seen, you know, I've been spending a lot of time on different content and research

[00:13:52] pieces lately with some of the top AI folks in the network, and they have just figured out really

[00:13:58] novel ways to extract insight from semi-structured or unstructured data, which is the case inside of

[00:14:05] most companies. Whereas figuring that out the first time took them months, right? But now they can

[00:14:11] apply this novel approach they developed and apply it as almost like a ready-made template or building

[00:14:17] block to the development cycle and now do that in two or three weeks. And that's the thing that's

[00:14:22] been most mind blowing to me when I spend time really going deep with them and try and understand,

[00:14:27] you know, all of this better for our own content, for our own marketing is what a colossal advantage

[00:14:33] that becomes. The frameworks that you guys use, I guess I was curious about like,

[00:14:38] do you have partnerships with some of these like platforms that may connect some of the, you know,

[00:14:46] agents or conversational AI, you know, bots and stuff like that? Or do you guys build like

[00:14:51] proprietary sort of plumbing? Yeah. I mean, we have some partnerships in place,

[00:14:54] like for instance, with, you know, writer.com. I'm familiar.

[00:14:58] May Habib's company. Okay.

[00:14:59] Yeah. So they're building basically the end-to-end platform for generative AI. But while the idea of

[00:15:06] AI is writer studio, which I think is very, very cool. And I'm a huge fan of May and her approach.

[00:15:12] I think she's one of the most, the smartest and strongest, you know, leaders in this space

[00:15:17] is to basically build no code tools internally for building AI applications. But the truth is inside

[00:15:22] of a lot of companies, you need more custom development that involves actually using,

[00:15:27] you know, the APIs with various tools to build it out. So having, you know, sort of writer certified

[00:15:33] A-teamers who can come in and execute against that is one example. And we're working to develop more

[00:15:39] and more of those relationships in the ecosystem because, you know, the truth is that every company

[00:15:44] is going to need custom development and most companies can't afford to go to an Accenture or

[00:15:51] McKinsey, you know, and it might also not always be necessary to do so. If what you really need is

[00:15:58] just a small agile tiger team to come in and help drive your development forward.

[00:16:04] Bob Pulver, host Q Podcast. Human centric AI, AI driven transformation, hiring for skills and potential,

[00:16:12] dynamic workforce ecosystems, responsible innovation. These are some of the themes my

[00:16:17] expert guests and I chat about, and we certainly geek out on the details. Nothing too technical.

[00:16:22] I hope you check it out.

[00:16:24] Excellent. I want to talk, you know, more about, you know, some of the things that you're working on.

[00:16:28] And personally, I know your newsletter, which is like has incredible, you know, following already,

[00:16:34] I think you hit 150,000 subscribers, at least on LinkedIn. But congrats on the success of that.

[00:16:40] And I thought you could just unpack that because I'm really liking your candid, you know, writing.

[00:16:46] It just resonates with me quite a bit.

[00:16:48] That's really kind of you to say, I really appreciate it. Yeah. So,

[00:16:51] you know, I've, I started out my career as a bit of a freelancer, you know,

[00:16:56] I started a media company out of college bootstrapped called the faster times of a couple of my bosses

[00:17:02] from college, but often to, to pay the rent and the bills in there. I was picking up a lot of

[00:17:09] freelance journalism gigs and content strategy gigs. You know, we built a branded content studio

[00:17:15] there doing early work with brands before content marketing was maybe even a term at the time back

[00:17:20] in 2010, 2011. And then spent the last 11 years working full time at platforms that basically

[00:17:26] specialized in connecting highly skilled freelancers who are at the top of their game with brands

[00:17:34] willing to pay at the top of the market for high quality, whether that was in, you know, content,

[00:17:37] filmmaking, now product development.

[00:17:40] So it got to the point where, you know, my creative side projects were getting to a pace where

[00:17:45] I thought, huh, I've been marketing and evangelizing this lifestyle for a decade now. Maybe I should get

[00:17:51] back to it a little bit because yeah, I, you know, I wrote a pilot of a TV show that got made with my

[00:18:00] friend or any partner, Shane Snow called Resignation. It's about a group of freelancers trying to make it in

[00:18:05] the new world of work. So we're currently, you know, hoping to get that that sold on a continuation deal.

[00:18:11] So I had that in the works was was helping out with some other TV shows that his studio is developing.

[00:18:16] You know, my newsletter, yeah, has continued to take off. So it's called the storytelling edge.

[00:18:21] It's about how to master the art and science of storytelling and use that to those lessons to

[00:18:27] build stronger relationships in your career and in your life. It's based off a book that I wrote

[00:18:33] that did pretty well a few years ago, also called the storytelling edge. And as AI hit,

[00:18:39] a lot of the big focus that I've had is really exploring what are the skills that are going to

[00:18:45] be most important in the age of AI. And as I dug into this for a lot of my future of work research

[00:18:51] at a team, the conclusion that I came to was that storytelling would actually be the super skill

[00:18:58] for this next stage. Because as we moved into an era in which machines could do the technical tasks

[00:19:04] and skills better than the vast majority of us, the thing that's going to be most valuable at work

[00:19:10] is the relations we have with other people, our ability to rally folks together, to communicate

[00:19:16] a clear vision, to set that path forward. And these are all the skills that we've long derided as soft

[00:19:24] skills, right? Things like empathy, communication, leadership. And if you really look at the course

[00:19:31] of human history of the way that these skills develop throughout our lives, the way that our

[00:19:38] brains work at a fundamental level, storytelling is the super skill that powers all of these different

[00:19:45] soft skills. So I've been writing a lot about how to really think about the role that storytelling

[00:19:53] and your own creative work plays in this next age of AI and what the future of storytelling looks like

[00:20:01] in this next era that we're entering into. So that's essentially the basis for this newsletter,

[00:20:08] which I'm investing more and more time in. As you alluded to, it has over 150,000 subscribers on LinkedIn,

[00:20:13] which goes direct to inbox. But I'm always honestly a little bit skeptical of a social media platform

[00:20:20] to do right by creators over the long term. So I've been migrating those folks for a while over to

[00:20:27] a kind of monthly mailing list and now moving that fully over to Substack and invest more time there.

[00:20:34] And that separate list has about 40,000 folks or so. And I'm working on a new book that's centered

[00:20:42] around a lot of these ideas, working on the TV projects and still involved with A-Team as a fractional

[00:20:49] exec, focused on a lot of our content and storytelling and PR efforts. But it's been two months now,

[00:20:56] two months today that I've been in this new role. And I have to say, I love it. I think as someone who

[00:21:02] always bites off a little bit more than they can chew, always has a bunch of side hustles going on.

[00:21:08] In addition to trying to raise a two-year-old, it's been nice to more formalize that as a portfolio of

[00:21:16] work, as opposed to trying to cram all those things in late at nights and on weekends and not sleeping

[00:21:23] enough. Yeah, for sure. I'm with you 100% on what I now call the durable skills. I give my friend

[00:21:30] Antonia Menaceri credit for that term. I know other people are starting to use it, but they really are the

[00:21:36] more durable skills going forward as you alluded. As AI takes on more and more, I was just having

[00:21:44] this conversation recently with some other guests and it's come up in a couple of different contexts

[00:21:49] about people's ability to tell, oh, I can tell when this is written by AI. Well, I know you think

[00:21:58] that, but let's be careful, especially if you're actually judging someone for college admission

[00:22:04] or for a job based on that writing sample, as opposed to someone that's, yeah, some people are writing,

[00:22:12] using AI to write their newsletters and stuff like that. And if you need to do that, that's fine,

[00:22:19] but you need to understand that if you're not the human in your own loop and content production before

[00:22:24] you hit publish, there's going to be plenty of people who just find that it's sort of dull. You

[00:22:29] just talked about this in your post the other day, right? Like if you think you're being authentic and

[00:22:35] expressing real human emotion and empathy and creativity, and you're talking about how you truly

[00:22:41] like resolved conflict or you've led an organization or whatever. I mean, you can sort of tell, but

[00:22:49] I just, again, I do caution people before judging too quickly because we do have false positives with

[00:22:57] some of this with people who maybe aren't, English isn't their first language or just people who need

[00:23:03] that thought partner and maybe just didn't do enough editing or whatever. But I think the points

[00:23:09] you made in that post, as I mentioned, they just really resonated. Yeah. I mean, I think one on the AI

[00:23:16] detection side, I think we have to accept that we do not have the tools yet to detect accurately when

[00:23:22] something is written by AI. So, you know, I heard a lot of people saying that, you know, professors

[00:23:28] that are using AI to check students' work and automatically failing them if they get a positive.

[00:23:34] I mean, AI tools will tell you that, these AI detection tools will tell you that the constitution

[00:23:38] was written by AI. Yeah.

[00:23:40] We're not there yet. We need to get there because we're seeing a deluge of AI slop pollute the web

[00:23:48] and things are getting really messy really fast. So we need better watermarking. We need better

[00:23:53] AI detection. Google just came out with a paper last week that seems to have found a fairly

[00:23:59] novel approach. I won't go deep into it because I don't fully understand it, but it essentially uses

[00:24:05] like a, a tournament method of checking AI against each other. That seems to be a decent solve.

[00:24:12] But when it comes to AI to create content, like it's one of my favorite topics to discuss right now,

[00:24:17] because I'm not against using AI in production or creative process at all. But where I think AI

[00:24:22] gets dangerous is when you outsource your thinking to it because writing is thinking,

[00:24:28] right? When writing is hard, it's because we're really trying to figure out

[00:24:31] what we want to say. So even when you go to AI for that first outline, you're outsourcing that

[00:24:39] thinking to the AI and whatever the output you're going to get from chat, GP to your cloud is really

[00:24:44] going to anchor your ideas. Inherently, the way that these frontier models work, both through the way

[00:24:50] that they're essentially an averaging out of all the texts on the web that has been scraped,

[00:24:55] copyright be damned in their training data, and also move towards safe outputs through the human

[00:25:01] reinforcement learning in, you know, essentially digital sweatshops in Africa and Southeast Asia.

[00:25:08] That all sucks the ideas and output into what I call this sort of vortex of mid or the ideas, you know,

[00:25:15] even with the best prompting, aren't going to be truly, truly original. They're going to be completely

[00:25:20] unoriginal, but they're going to be just kind of mid. And, you know, the quality, it's going to be better than,

[00:25:26] you know, even a lot of freelance writers, you might be able to hire certainly better than an intern.

[00:25:31] But it's a little bit dull and soulless. One of the best jokes I've heard is, you know, that AI art and

[00:25:38] writing has done what humans can ever do and proven the existence of the soul just by reading it and

[00:25:45] seeing the lack of it. But, you know, if you do really start by figuring out what are your own

[00:25:52] unique original ideas, if you do hook your audience with stories about your own life, because AI can

[00:25:58] never come up with stories about your own life for you, right? AI might be able to write, but it can

[00:26:02] never be a storyteller. If you do those things, and then you use AI to get feedback on your writing,

[00:26:08] to clean it up, to give you like alternate tones for specific paragraphs, use things like Grammarly Pro

[00:26:14] for, you know, copy editing, like, have that. Like, I think all those use cases are totally fine.

[00:26:20] But I do think that people are going to gravitate towards content that they feel like is created by

[00:26:27] a human. And that's what those stories and even some of those imperfections in human created content

[00:26:32] give you. Like, is your LinkedIn feed like mine flooded with people making those avatar videos

[00:26:39] and being like, this is crazy, this is going to change everything, you know, this was a fake video

[00:26:43] of me? Yeah, I've got a lot. And I've also talked to some folks who work at some of those, those

[00:26:49] companies who are making those, those avatars and have offered to use that. But my very last

[00:26:56] conversation was about, like, whether or not that's an appropriate use of the technology. Like, if you

[00:27:03] were, let's say you're not great at public speaking, and you just want to get something out, you want

[00:27:08] it to be as polished as it can be. Do you create a digital twin feed it a script, and then just have

[00:27:16] this like super polished, you know, almost magical, like version of yourself? I mean, don't we do that

[00:27:23] with, if we're using AI to create a custom, you know, resume, so we can put our best foot forward? I

[00:27:28] don't know. I don't know where the line is drawn. I just don't know who wants that, right? I think

[00:27:33] it's a nobody wants this scenario. Yeah, the only type of content that works with those digital twins

[00:27:40] is the ones pointing out that this is novel, like, hey, this is, I created this video of my digital twin

[00:27:46] that's talking to you right now. Isn't this crazy, right? That's the one type of content with that,

[00:27:51] that works. I don't think anyone is going to follow a creator who's constantly creating uncanny

[00:27:59] valley content using their digital twin over and over again. Like you're just sort of losing that

[00:28:04] sense of trust and authenticity. The reason that vertical video has popped off so much and why,

[00:28:12] most people under the age of 25 spend way more time watching, you know, vertical video on TikTok and

[00:28:19] Instagram than more premium polished, you know, TV and video is because we're drawn to the authenticity

[00:28:28] of that format. Having an AI version of yourself doing those vertical videos completely defeats the

[00:28:35] point. And I think that's the thing that a lot of people in Silicon Valley just don't seem to get about

[00:28:41] the way that humans operate in terms of content is that we just don't necessarily want a lot of these

[00:28:48] things when it comes down to it. One of the dumbest ideas I feel like I've heard is the idea that we're

[00:28:54] all going to have personalized versions of our, of TV shows, uh, where I'm getting a version of,

[00:29:00] you know, Game of Thrones or House of Dragons. It's only made for me. Nobody wants that TV and stories are

[00:29:05] a communal experience, right? Like it would suck to see an episode of a TV show that I can never talk to

[00:29:12] anyone else about because they never experienced that same episode. It, there's just all these

[00:29:17] little things in the way that we think about AI and content that I think completely missed the point

[00:29:22] of how humans operate and what we're actually attracted to. Yeah. I think that's taking.

[00:29:27] Peter Zollman

[00:29:27] Hi there. I'm Peter Zollman. I'm a cohost of the Inside Job Boards and Recruitment Marketplaces podcast.

[00:29:34] And I'm Steven Rothberg. And I guess that makes me the other cohost.

[00:29:37] Every other week we're joined by guests from the world's leading job sites.

[00:29:41] Together, we analyze news about general niche and aggregator job board and

[00:29:46] recruitment marketplaces sites. Make sure you sign up and subscribe today.

[00:29:52] The choose your own adventure concept just a little bit too.

[00:29:56] Yeah. I mean, like, listen, you know, like I read some choose your own adventures

[00:29:59] goosebumps when I was a kid, like I enjoyed them. But, uh, even with that, like you're getting the same

[00:30:04] choose your own adventures as everyone else. Right. I can still talk to, you know, my best friend in,

[00:30:09] in first grade about like, oh, did you try, you know, number seven. And then, uh, it's still a

[00:30:14] common experience. I just don't think anyone wants it.

[00:30:17] I'm thinking about what you said about like what the younger generation really pays attention to.

[00:30:23] So, uh, obviously kids are spending too much time on, on social media and they're doing scrolling and

[00:30:28] they could probably be doing something more constructive. But to your point about the content itself,

[00:30:32] I mean, my daughter, my daughter's 16 now, but I think some of the stuff that she watches is just

[00:30:38] incredibly sophomoric and stupid, but at least it's authentic. And at least it's real people

[00:30:43] doing stupid shit, but you should maybe appreciate how people built those things and how,

[00:30:50] I don't know how it works or whatever.

[00:30:52] David Morgan Everyone's always wanting an easy button

[00:30:54] for content, right? So digital twin of yourself, where you don't have to get good at delivering your

[00:30:59] own video and you can just have AI do it for you. Like that's an easy button, but I think in the end,

[00:31:04] it's not really going to work.

[00:31:05] Yeah.

[00:31:05] David Morgan In terms of what AI formats, you know, I think that's just the stupid idea. The ones that

[00:31:10] scare me are the character bots right now, especially from a parent perspective. Like,

[00:31:15] imagine you heard of that, you know, that kid and, um, who killed himself after falling in love

[00:31:20] with his character AI bot and wanting to, you know, be with her. Um, and I think that's the type of stuff

[00:31:26] that's really scary moving forward.

[00:31:28] David Morgan Yeah, for sure. And I mean,

[00:31:30] in some ways that's, um, you know, when we talk about responsible AI, I mean, that's just one of

[00:31:36] those things that we've got to really think about. I've been more focused obviously on organizational

[00:31:42] use cases, especially in talent where we're making decisions about people's livelihoods, right. But,

[00:31:49] you know, because, you know, we have, we have kids and because we see what's going on

[00:31:54] elsewhere in other domains. Um, we've got to think more broadly about, you know, responsible

[00:31:59] use and being responsible by design and having the right controls, you know, in place as well as

[00:32:04] the transparency to know AI is being used. So I think I definitely want to check out what Google

[00:32:11] just published. I saw that, um, as well about the watermarking.

[00:32:13] David Morgan Yeah. Just on the, on the content, uh, side and the story, the storytelling

[00:32:21] being a little, you know, selfish, but I'm sure there's listeners that are struggling with the

[00:32:25] same thing, but I know that there's ways that I can use AI to be my, my thought partner. But

[00:32:32] to your point before, like, I don't want it coming up with, you know, random

[00:32:37] shit that I try to like, you know, expand upon. I want it to learn enough about me and some of the,

[00:32:42] some of the ideas that I have and some of the things that I've done. So it at least knows,

[00:32:47] you know, how to sort of, uh, assess what I do right and, and strengthen it. But any advice for

[00:32:54] someone who's done and seen a lot of shit and just has trouble, uh, getting it, getting it out there.

[00:33:01] David Morgan Yeah. I mean, let's do a little

[00:33:02] writing therapy right now. Like, what do you find most difficult? Is it like figuring out,

[00:33:07] you know, what that, that hook is for your post? Like what's a story to lead with? Like,

[00:33:12] where do you find it most difficult or just the act of writing itself, like sitting down and doing it?

[00:33:17] David Morgan More of the latter,

[00:33:17] but a lot of it is like my brain kind of thinks of many things and the connection between those things

[00:33:23] all at once. And so it's really just like the focus and the logic of laying out like,

[00:33:29] David Morgan And prioritizing maybe what's important. It's just my, you know,

[00:33:34] ADD brain trying to get it down. I mean, you know, my editor always has a lot of work to do because

[00:33:40] even through these conversations, I kind of go in different directions, even stopping mid-sentence

[00:33:46] sometimes and, and thinking of something else. Right. So it's like, I don't want to spend,

[00:33:50] shouldn't take me, you know, 10 hours to write a couple paragraphs. Right. So it's just,

[00:33:54] it's just getting it out there. And maybe there's a way to just have like a, like a repository of,

[00:34:00] of all of these thoughts. Maybe I just need to start writing them down and putting them away and

[00:34:04] then giving AI access to it to say, let's figure out how we can formulate some of these key thoughts in

[00:34:10] a logical fashion. So it flows like a, like a story. Right. I think that is one of the interesting

[00:34:15] applications right now of the Google LM notebook. So that was developed at Google with this writer,

[00:34:25] Stephen Johnson's written, you know, a dozen books, maybe 14, you know, pretty popular books,

[00:34:30] New York times writer when in house at Google developed this tool and basically was tasked with

[00:34:35] like help us develop an AI tool that will actually be useful for writers. I was really skeptical of Google

[00:34:41] notebook when it came out, because I thought the automated podcast thing was just another example

[00:34:47] of like a hyped AI content format that would actually have very little utility in real life,

[00:34:53] especially because it makes things up. It changes ideas back to the most average version of them.

[00:34:58] It's a lot of garbage filler if you actually listen to a few of them. But I think what's really

[00:35:04] interesting about this tool is to have this space where they're not feeding your own content,

[00:35:09] a back into their own training data, which is important, but essentially where you could dump

[00:35:14] a lot of different ideas. And then you can talk to the notebook about connecting those different

[00:35:20] concepts together. Let me give an example. So I think that the simplest sort of formula for thought

[00:35:26] leadership writing is lead with a personal story about your life that hooks the reader in,

[00:35:34] and then that connects to a bigger idea that you might have, right? For example, I give a popular

[00:35:41] newsletter I wrote a few weeks ago about why storytelling will be the super skill of the AI age.

[00:35:47] Open with my experience, like having my son a few weeks after chat GPT comes out, spending my maternity

[00:35:55] leave going into an absolute doom loop rabbit hole over generative AI, worrying about his future,

[00:36:01] worrying about my future, and how that then led to these realizations about how storytelling actually

[00:36:07] will be the most important skill of the AI age, right? This is the sort of formula that just works

[00:36:12] really well for any sort of thought leadership content because it creates a relatable connection

[00:36:16] with the reader and then makes them more open to this big idea that you're presenting because they feel

[00:36:23] a connection to you, you've built some trust, and it brings them in as opposed to just starting with

[00:36:28] the idea which feels like you're doing homework or you're eating your vegetables off the bat.

[00:36:33] I think there's potentially interesting applications of using a notebook like that to connect sort of

[00:36:38] big ideas that you might have to personal stories that you might have because I think everyone does this

[00:36:44] in a different way. Like I have some writer friends who start with a little story and then they go and

[00:36:49] they're searching for what is an idea that could possibly connect to that story, right? As an opener,

[00:36:54] I'm someone more who has the idea that I want to write about and then I look back in the recesses of

[00:37:01] my brain for what is the story that I could connect in there. But that seems like the type of thing that

[00:37:06] AI could help if in your ADD brain, and I think one of my specialties as an editor is working with ADD

[00:37:13] writers if you ask some people on my team because I kind of can see the different disparate connections

[00:37:18] that they're making in their brain, but is to sort of like have two dumps, right? Like one notebook

[00:37:24] that's like all the kind of big ideas and concepts I want to write about. And then I have over here

[00:37:28] is like a repository of different story like having my kid like hitchhiking in Albania, jumping off this

[00:37:34] clip in Hawaii, getting arrested in Costa Rica for doing some bad shit. And then figuring out how you can

[00:37:44] tie those opening leads to your big ideas. I think that could be like one really interesting

[00:37:50] application you have. Or I'm also bullish on just using AI as a little bit of a tutor or a

[00:37:55] thought partner. I think voice mode is really cool with chat GPT. I've had surprising results just like

[00:38:02] walking around with my dog and talking to it and trying to work out an idea, right? Like,

[00:38:07] does this make sense? Like how would I refine it? Yada, yada, yada. Like you're not always going to have

[00:38:13] someone that you can talk to on demand about that. So treating, you know, AI like a AI intern that you

[00:38:20] can just brainstorm with, I think is another way that we can do that.

[00:38:24] No, that's really helpful. Yeah. I mean, some of it's like, I've done a lot of stuff and a lot of those prior

[00:38:31] experiences, even if they were quite a while ago, like they're coming back full circle, like another wave of

[00:38:39] technology, right? Like everything that I was describing at the beginning that relates to like

[00:38:45] A-team and putting that, putting the teams together and stuff like that. I mean, all this extracurricular

[00:38:51] work I was doing at IBM around like crowdsourcing and collective intelligence before there was a gig

[00:38:55] economy. And before any of this really took off, it's like, it was all manual. And then, and then IBM

[00:39:02] research came along and started giving us these, you know, graph database powered solutions for social

[00:39:06] network analysis and knowledge graphs and all this stuff. And it's like, it keeps resurfacing in the

[00:39:11] context of AI, like the, everything I went through with building automation strategy at NBC universal,

[00:39:16] or, you know, early, the first wave of, of chat bots that completely sucked. So there's all these topics

[00:39:22] that are become interrelated. And then I've got to like, put them in a logical fashion, if I'm going

[00:39:27] to incorporate them in a particular story at all. And then I guess the other aspect is, you know,

[00:39:33] I spent most of my career as an individual contributor, sometimes getting pretty deep

[00:39:37] in the weeds, but then other times I'm also like a chief of staff who's got this higher profile kind of,

[00:39:45] kind of audience. And I have no choice, but to balance both, both being in the weeds to do tactical

[00:39:52] stuff and get my hands dirty, as well as putting my systems thinking hat on and thinking big picture,

[00:39:57] like how is this going to evolve or how could this evolve over time on a H2, H, well, probably not

[00:40:05] H3, that's just throwing darts, but moving from H1, H2, like what happens when you get out of the

[00:40:10] trough of disillusionment, right? Like, so it was just always like 20 things going on in my head.

[00:40:16] And so, so maybe it's just the focus and the dedication and then a little bit of perfectionism.

[00:40:21] It kind of makes me really intrigued with the idea of building a sort of story editor or story teaser

[00:40:29] GPT like that, right? Because I sat with you, I could ask a series of questions that gets to the

[00:40:34] core idea you'd want to do. Be like, okay, like thinking back to your time at ABM, like you're

[00:40:40] building those chatbots. Like was there one moment where you realize like this just isn't going to work,

[00:40:44] right? And then could we connect that story to a development that you see right now where, you know,

[00:40:51] maybe it's Claude and the way and where we are with agents, right? Where, you know, there's a lot of hype

[00:40:57] around agents in this next era of work. But we saw earlier this year that Devin, the AI software developer,

[00:41:05] turned out to mostly be a scam. Even Claude's agent capabilities that came out two weeks ago,

[00:41:12] I mean, it can only actually operate an operating system successfully a little less than 15% of the

[00:41:18] time, you know, and I imagine you could, I could probably like program a GPT to do that somewhat

[00:41:24] successfully. So if that's something I'll try out, then I'll send to you to see if it helps at all.

[00:41:29] Yeah, that'd be awesome. Joe, I want to be respectful of your time. I know you've got a two-year-old

[00:41:34] waiting to get some candy. So any final thoughts on advice for folks that haven't gotten started to,

[00:41:41] you know, literally elevate their AIQ.

[00:41:44] So be curious. It's the most generic advice you're probably going to get. But it's the thing that

[00:41:49] Ethan Mullick writes about in his book, Co-Intelligence, which I highly recommend,

[00:41:54] by the way. I know we're not doing the Ezra Klein podcast thing of recommending three books at the

[00:41:58] end, but that's one that I'd recommend. He writes that you only really get an understanding of what

[00:42:03] AI tools are capable of when you spend at least 10 hours using them. And I think that's right.

[00:42:07] And that's in my experience with every tool that I've used, whether it's Claude, whether it's a

[00:42:11] video editing tool like Opus, whether it's like Mid Journey, but figure out where these things can

[00:42:16] play into, you know, your productivity and your efficacy in your work with AI. I think that from

[00:42:24] a storytelling perspective, like we're looking at three horizons right now for how, you know,

[00:42:30] content and storytelling is going to change in the age of AI. For my creatives out there, I think the one

[00:42:35] we're at right now is using it in terms of, you know, enhancing productivity, using it like your

[00:42:41] team of interns. You can build your one person media operation. You couldn't before using it for

[00:42:45] tutoring, doing the sort of teasing out of story ideas like we just talked about or learning or

[00:42:50] researching about a topic more deeply. But I also think we're getting into a second wave where we're

[00:42:54] seeing some really creative artists and filmmakers use AI in a way to create something that is new and

[00:43:02] original and artistic. And, you know, this stuff could go down a really evil pathway in terms of

[00:43:09] just taking all of our jobs and having the world polluted with slop, which definitely is happening

[00:43:14] already. But there's also a world in which, you know, we see short films like Airhead, which is made

[00:43:20] using Sora that are new and different and interesting and things that we couldn't create without it.

[00:43:24] That's the path that most technology takes in storytelling. And, you know, the third horizon that I'm

[00:43:29] most interested in researching right now, so if anyone listening to this has seen a cool example

[00:43:33] of it, shoot me a DM on LinkedIn. My DMs are open. It's using AI for sort of collective intelligence

[00:43:43] and co-creating art in big groups. You see artists, you know, like Grimes and Holly Herndon

[00:43:49] doing this in the music space. And yeah, finally, like if you want to follow this more deeply,

[00:43:55] if you want to learn about the art and science of storytelling and how AI is changing content,

[00:43:59] give a plug for, you know, my LinkedIn newsletter called the Storytelling Edge. You can also find

[00:44:04] on Substack, the Storytelling Edge, go to my profile, you can find all those things and

[00:44:10] appreciate the opportunity to get to chat about this stuff. Always, always love to, Bob, and

[00:44:15] the opportunity to fit in a couple of plugs here.

[00:44:17] Yeah, of course. Of course. Yeah, Joe, I really appreciate your time. This was awesome.

[00:44:21] Thanks for some of the personal tips as well. And yeah, I want to wish you the best of luck with the

[00:44:27] new book and the newsletter and the movie and all these other things that you're working on. And

[00:44:34] I hope your son gets a lot of good candy. Appreciate it. Yeah, probably going to limit him to one or

[00:44:38] two pieces, but you know, I'll eat the rest. Nice. So yeah, and some of those links I'll make sure to

[00:44:45] include in the show notes so people have them. Thank you again, Joe. This was great. Thanks everyone for

[00:44:50] listening and that concludes another episode of Elevate Your AIQ. We'll see you next time.